review article a review study on traction energy saving of...

Hindawi Publishing CorporationDiscrete Dynamics in Nature and SocietyVolume 2013, Article ID 156548, 9 pageshttp://dx.doi.org/10.1155/2013/156548

Review ArticleA Review Study on Traction Energy Saving of Rail Transport

Xuesong Feng, Hanxiao Zhang, Yong Ding, Zhili Liu, Hongqin Peng, and Bin Xu

MOE Key Laboratory for Urban Transportation Complex Systems Theory and Technology, Beijing Jiaotong University,No. 3 Shangyuancun, Haidian District, Beijing 100044, China

Correspondence should be addressed to Xuesong Feng; [email protected]

Received 4 July 2013; Revised 3 August 2013; Accepted 21 August 2013

Academic Editor: Wuhong Wang

Copyright © 2013 Xuesong Feng et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The energy cost of a rail transport system is very big especially for the tractions of trains in its daily operation. Continual effort ondecreasing the traction energy cost intensities of various rail transport modes has beenmade bymany researchers and practitionerson different aspects for a long time. From the rail transport operation perspective, this study reviews such energy-saving researchmainly focusing on the optimizations of train coasting schemes, the rational designs and utilizations of track alignments, theameliorations of train attributes, and the improvements of system operations. According to the review work, it is confirmed that insound responses to distinct track alignments, dynamically optimizing control programs of trains with their reasonably improvedattributes ought to be further studied in view of the systematic transport operation of a rail line or network in an integrated manneras much as possible in the future research on traction energy saving of rail transport.

1. Introduction

Although rail is commonly regarded as an energy-savingtravel pattern, the operation of a rail transport systemconsumes in fact a huge amount of energy every day. Forexample, theweekly electric cost ofmerely one railway stationof Hong Kong Mass Transit Railway Corporation reaches230 megawatt hours [1]. Such huge energy consumptionsometimesmay have a serious energy waste. For instance, it isastonishing that the energy cost (EC) per passenger trip of atransport completed by themetro system inNewYork is evenmuch higher than the EC of the same trip by car on average,mainly due to the low utilizations of the passenger capacitiesof the metro trains [2].

Therefore, only rationally utilizing a rail transport modeis able to avoid unnecessary EC to achieve its sustainableoperation and development. For example, optimizing thestreamline design of a train [3], increasing its passengercapacity utilization rate to some reasonable levels [4, 5],coasting the train, that is, in other words, taking advantage ofits inertia motion, as much as possible in its transport process[5, 6], and so on all in practice take much positive effect ondecreasing its EC per unit transport. In addition, tractionpower innovations are also able to not only effectually

improve the EC efficiency of a train [7] but also substantiallyreduce its emissions of NO

𝑥, CO, PM, and so forth [8].

2. Energy Cost Factors

The EC of a rail transport system for its daily operationconsists of two parts, that is, the traction EC (TEC) for thetractions of its trains and the additional EC for, for example,lightening in stations, repairing cars in depot(s), and so forth.The share of the TEC is ordinarily muchmore than half of thetotal EC of a rail transport system. As shown in Figure 1, theTEC intensity (i.e., the TEC per unit transport) of a train forits completion of a transport task is dynamically determinedby many factors on the aspects of driving tactics (i.e., mainlythe coasting program) of the train, alignments (e.g., theslope gradients) of the track, attributes (e.g., the start-stopfrequency) of the train, operationmanagement (e.g., the trainscheduling) of the rail transport system, natural environment(e.g., the wind [9]) and others (e.g., the aerodynamic forcesespecially in a tunnel [10]) in an integrated manner.

Focusing on the impacts of various factors, manyresearchers and practitioners have been continually for a longtime making effort for the decrease of the TEC intensity ofa train or, afterwards in a further, multitrains operating on

2 Discrete Dynamics in Nature and Society

TEC intensity

Driving tacticsTrain attributes

System operation

Track alignments

Nature environment and others

· · ·

Figure 1: Influence on TEC intensity.

a rail transport line or network and have proposed manyvaluable suggestions. This study reviews these works fromthe transport operation perspective mainly with respect totrain coasting control optimization, reasonable track align-ment design and utilization, train attribute amelioration, andrail transport system operation management improvement.Accordingly, the direction of our effort in future research isindicated for radically decreasing the TEC intensity of the railtransport system operation in an effective way.

3. Coasting Control

A train consumes energy in a relatively very slight intensitywhen it coasts. Rational site choices of a train for the startsof its coasting along a rail line are able to effectively decreasethe TEC of the train for its whole trip by avoiding its brakesfor changing its speed with much loss of its kinetic energy. Asillustrated in Figure 2, much importance in research has beenattached to the optimization of the coasting scheme of a trainfor its energy saving in regard to transport time expenditure(TTE), train traction performance, passenger ride comfort,and so forth.

Based on train traction calculations [11, 12], a transportsimulation program following the principle of coasting vector(i.e., coasting speed and coasting start point) to achieve theoptimum automatic train operation with the least TEC fora trip in a certain travel time is developed by Chui et al.[13] for Kowloon-Canton Railway (KCR). A train is given acoasting vector data from the trackside so that whenever thetrain has actually reached the coasting speeds or coasting startpoints determined with a view to passenger travel demandsin different daily operation time of a rail line, its motoringpower will be shut off. In this way, about 3% of the TEC couldbe saved for KCR.

For the sake of improving the application result veracity ofa train coasting program, Chang and Sim [14] newly proposea genetic algorithm (GA) [15] applied in their rail transportsimulation study which takes moving block signals [16] toinsure the safe distance between two neighboring trains run-ning on the same rail line. Different sites of a transport sectionbetween neighboring stops and various control actions of atrain are respectively regarded as genes and chromosomesin the GA to search for the optimal coasting start pointsof the train. Decreasing TEC, ensuring on-time running,and avoiding jerky actions (i.e., in other words, providingcomfortable transport service to the passengers) of a train are

TEC intensityCoasting optimization

Traction performance Etc.Ride comfort TTE

Figure 2: Coasting optimization for decreasing TEC intensity.

considered in a comprehensive way to optimize its drivingtactics for a transport mission.

Because of the computational complexities of GAs for theexploration of the optimal coasting program in comparisonwith, for example, gradient algorithms [17] especially whenthe stop spacing of a train is comparatively very short[18], Hwang [19] tries to simplify the process of optimizingcoasting controls of a high-speed railway (HSR) train for itsenergy saving.The relationships between the economic speedof a HSR train and its TEC and TTE for a passenger transportwork are built based on fuzzy clustering analyses [20, 21]. As aresult, it is possible to get the range of the economic speed ofa train according to expected TEC and TTE. Thereafter, thecoasting program of a train is resolved by taking advantage ofa hybrid scheme of GAs to optimize the speeds of the train atdifferent rail sites within its economic speed range.

Due to low efficiencies of classic numerical optimizationmethods [22] and burdensome train traction calculations indetailed simulations of the transport controls of a train [5],it is difficult to truly realize the optimum on-board controlsfor the minimal TEC. On the basis of existing studies, Bai etal. [23] develop a set of train operation guidance equipmentsto successively give suggestions to the driver of a train inits transport process for optimizing the driving actions fortraction energy saving to some dynamically real-time extent.Liu and Golovitcher [22] provide an analytical approachsolved by maximum principles [24, 25] for relatively efficientdecisions of optimal control change points of a train adoptingfull or partial braking, coasting, or motoring with partial orfull power. Moreover, according to the statistical distributionof the operating conditions described by the speeds andaccelerations of a train on the macrolevel, Lindgreen andSorenson [5] simplify the transport simulation work for theversatile and flexible control strategy (including coasting startpoints) optimization with an acceptable TEC computationaccuracy.

Based on previous research on the effect of optimizingthe coasting plan of a train on its TEC intensity, analyses of

Discrete Dynamics in Nature and Society 3

the additional impacts of traction acceleration and brakingcapacity of a train are made by Bocharnikov et al. [26] insimulation with utilizing dynamics theories [27], GAs, andfuzzy mathematics (FM) [28]. Different coasting programsof a train for the same transport work are comparativelystudied with focusing on the changes of its TTE as well asTEC. The coasting scheme is finally determined to preventthe unilateral optimization of the TEC at overmuch expenseof the TTE. The utilized GA is able to effectively avoidimproperly adopting local optimal solutions in searching forthe most energy-saving coasting start points for the wholetrip. Moreover, variables applied in the dynamic simulationsare well defined by FMmethods.

With regard to transport time schedule adherence, KimandChien [29] newly develop a train performance simulationapproach to optimize the controls of a train for distincttypes of track alignments to decrease its TEC for a certaintrip. Under some transport time constraints for a transportdistance between neighboring stops, the most energy-savingcontrol scheme of a train for the transport section is deter-mined according to the optimal selection of its coasting startpoints. In this respect, simulated annealing approach [30] isutilized to analyze the TECs of different operations of thetrain for various track alignments in the transport sectionwhich is divided into many small subsections. Further, Kimand Chien [31] improve their work with the additionalconsideration of the formation of a train. The sensitivityanalyses of different factors such as track alignment, speedlimit and train formation of a train are also made for itsenergy saving.

4. Track Alignments

As previously-clarified from the viewpoint of optimizing thecoasting scheme of a train, various types of track alignmentshave different influence upon the TEC intensity of the sametrain control action. Therefore, by effectively transformingpotential energy of ramps along a rail line [32–34] andincreasing running straightness of a train in plane [35],reasonable designs and/or utilizations of the track alignmentsare much beneficial to reducing the TEC intensity of a train.Valuable studies have beenmade on this aspect in view of, forexample, driving strategy of a train, headway of trains, andride comfort of passengers, as interpreted in Figure 3.

Focusing on updating the ramps in stations along ametroline for different transport distances between neighboringstops, Hoang et al. [36] first attempt to simultaneously opti-mize vertical track alignments and train operation controlfor traction energy saving. Different patterns of the rampsare defined according to realities. Heuristic algorithms [37]and dynamic programming (DP) [38] are used to detect theslope change points of each type of the ramps. Thereafter,a simulation approach is applied in accordance with theprinciples of train traction calculations to compute the TECof a train for its trip from one terminal station to another of arail line.The designs of the energy-saving slopes of the rampsin each of the stations and the corresponding actions of a trainon different rail sites are able to be optimized rationally.

TEC intensity

Ride comfortDriving actions Etc.

Track alignments

Headway

Figure 3: Track alignment design and utilization for reducing TECwaste.

From a unique viewpoint, Firpo and Savio [39] doresearch on how to reduce the electric energy provided bya transformer substation to a transport section of an HSRline for the transports of a certain number of trains withinsome time. The impacts of the track alignments togetherwith the site of the transformer substation along the railline and the headway of the trains in the transport sectionon the electricity supplied by the transformer substation arestudied comprehensively. With the help of electromechanicalsystem modeling approach [40], the established nonlinearoptimization model is solved by both Simplex method [41]and stochastic hill-climbing algorithm [42, 43] to explore theoptimal combination of target speed, acceleration, and timein stop of a train for the transport section.

On the assumption that the climbing ability of a metrotrain is capable of coping with all kinds of ramps, Kim andSchonfeld [44] study simulation of the influence of varioustypes of vertical track alignments upon the intensities of bothTEC andTTEof the train in consideration of the ride comfortrequirements of the passengers. It is concluded that if a stopspacing is shorter than about 2,400 meters, a rational slopedesign of a ramp in station is able to take obviously positiveeffect on saving the TECof a train for this stop spacingmainlybecause of its coasting after the ramp.Moreover, reducing theacceleration of a train is of benefit to decrease its TEC forthe same transport work, but meanwhile, its TTE increasesin correspondence.

Regarding impacts of different radians of various typesof track curves in plane on the TEC intensity of a train, Liuet al. [35] make simulation analyses with employing traintraction calculations. It is demonstrated that the increaseof the TEC intensity of a metro train with the decrease ofthe track curve radius is obviously accelerated when thecurve radius becomes smaller than about 300 meters. Incontrast, if radiuses of track curves are over approximately500 meters, their radians have almost no influence upon theTEC intensity of a metro train.

It is true that the speed adjustment action of the driver islagged behind the change of the speed (control reminding)code of a train when the track gradient, speed restriction,and so forth are various between successive stops. As a result,Ke et al. [45] propose a combinatorial optimization modelto minimize the TECs of a train for each of its transportsections. A MAX-MIN ant system [46] is used to optimizethe train-speed trajectory for reducing the computationalburden on the block-layout design. Further, the optimalspeed codes of a train are also determinable for its transport indifferent sections to provide an energy-saving, efficient, and


comfortable service [47]. In order to insure the practicabilityof corresponding driving strategies, the relationships betweenthe acceleration and speed of the train and the track gradientare regulated with a fuzzy process [48, 49] to reflect thecorrelations among track alignments, train actions, andcorresponding TECs.

5. Train Attributes

Besides the optimum driving controls of a train and therational designs and utilizations of track alignments, ame-liorated attributes of a train also play important roles inits traction energy saving. Different from the viewpoints ofaforementioned studies, many works pay much attention tofor example, motor flux, weight, mass distribution, electricitysupply effectiveness, traction power output, startup-stopfrequency, formation, carrying capacity utilization rate, andso forth of a train for the improvement of its TEC efficiency.As shown in Figure 4, such research involves differentfields including electrical control, rail car manufacturing, railtransport operation management, mechanics, and electricpower supply.

From the perspective of regenerative braking of a train,Kokotovic and Singh [50] as early as in the 1970s haveproposed an optimization approach to provide the basis forthe improvement of train control circuit through optimizingthe electric motor flux control to reduce the energy costintensity of the traction motor of a train. With referringto the correlations between the speeds and traction forcesof a train for different track alignments [11], the energy-saving transport objective of the train is achieved by utilizinga Hamiltonian system [51] with adjoint analyses [52] torationally adjust the speed of the train through tractionmotorflux control. In addition, the relationship between the TECand TTE of the train is also studied with the application ofGreen’s theorem [53] for further optimization of the motorflux control.

According to train traction calculations applied in sim-ulations of the transport procedures of different types ofChinesemetro trains, it is empirically confirmed that theTECintensity of a metro train in China is increased with raisingits weight in an approximately linear way [35]. Reducing theweight of a train is a directmeans to decrease its TEC intensitythrough reducing the resistances to the train in its trip. On thebasis of the detailed survey on various technical specificationsof different types of trains inGermany, IFEU [54] analyzes thegeneral energy-saving effect of reducing 10% of the weight ofeach type of the trains. It is revealed that the vast majorityof the TEC of an urban or regional rail transit train for itswhole trip is used to accelerate the train in its startups becauseof its comparatively very short stop spacing on average. Asa result, if the weight of such a train decreases by 10%, it isable to save about 7% or 8% of its TEC for a trip in Germany.In comparison, an intercity train with a relatively long stopspacing needs to spend roughly half of its TEC in overcomingthe air resistance to its moving after a startup, and therefore,the TEC saving by lighting the weight of the train is from 4%to 5%.

TEC intensity

Transport management(stop frequency, etc.)

Train attributes

Mechanics(engine power output, etc.)

Manufacturing(weight, etc.)

Electrical control(motor flux, etc.)

Power supply(electricity supply effectiveness, etc.) Etc.

Figure 4: Attribute improvement for TEC intensity reduction.

As for simulation research on rail transports, initially amass point or afterward a line with itsmass evenly distributedis commonly used to roughly represent a train. In contrast,Chou and Xia [55] analyze the impact of the mass differenceof various parts of a freight train on the TEC and TTEof its whole trip with regard to transport security. In theirsimulation work, a heavy haul train is divided into a certainnumber of units according to the traction power distributionof the train. Every two neighboring units is separated withprotective equipments controlled by electric signals in anassociated way to dynamically adjust actions of differentunits confronted with different track alignments. Optimalcontrol strategies to reduce the TEC and TTE of a transportand improve its security are suggested on the basis of thesimulation analyses.

For the purpose of decreasing the TEC intensity of adiesel multiple-unit train, Lu et al. [56] evaluate differentcombinations of the traction power outputs from the multi-engines of the train for a certain total traction power output.DP has been applied to analyze the TECs produced by asingle-train motion simulator to identify the global optimalprogram of the traction power distribution between differentengines of a train for a certain transport whose profilemust be obtained in advance. To realize the online powermanagement [57], an adaptive online strategy for the real-time optimization based on the results from the DP studyis put forward. It is indicated that the optimized solutionhas a TEC reduction of around 7% in comparison with evendistribution of the total traction power output of a train toeach of its engines for the same transport mission.

A high startup stop or accelerating-braking frequency ofa train substantially increases its TEC for the same transporttask [5, 6]. For example, an express train running without anystops from Copenhagen to Elsinore is able to save about 50%of its TEC in comparison with its transport with every stopfor this trip [5]. Feng [58] and Feng et al. [59], respectively,prove that the increase of the TEC intensity of an HSR ormetro train with shortening the stop spacing is apparentlyaccelerated when the stop spacing becomes shorter thancorrespondingly around 100 kilometers or 1,800 meters, andsuch a trend becomes more obvious if the target speed of thetrain is improved. In consideration of short distances betweenneighboring stations ofmostmetro lines in reality, a relativelylow target speed of a metro train [35, 59] and its frequentcoasting operations are able to effectively decrease its TECintensity [6, 60] especially for relatively low boarding rates inoff-peak hours when the TEC intensity of the metro system


is much further decreased if the trains extend their headways[13].

As widely recognized, it is comparatively easier to makethe utilization rates of the passenger capacities of shortlyformed trains relatively high according to dynamicallychanged travel demands of the passengers, which will obvi-ously decrease the total TEC intensity of the trains running ona rail line or network [31]. Moreover, the decrease of tractionpower of a train is able to reduce its TEC intensity with moreTTE [26, 44]. In addition, Feng et al. [61] also reveal thatin comparison with the TEC of a short HSR train with thesame equipments, an HSR train with a long formation mayhavemore TEC for the same trip due to the speed restrictionsof the exit switches of the tracks in stations. All these factsrequire systematically reasonable and flexible applications oftrains with different formation lengths and traction capacitiesin view of detailed passenger travel demands in differentoperation times or seasons of a rail transport system.

6. System Operation

Thesustainable as well as efficient operation of a rail transportsystem is supported by not only superior performance ofits equipment and facilities but also rational system oper-ation management. However, because of, for example, spa-tially uneven transport demands as a whole, only aggregatestatistics on the overall TEC intensity is unable to reflectthe demand variety at different parts of a rail transportsystem [62] and certainly hard to provide adequately effectiveenergy-saving approaches in systematic concepts. Takinginto account interactions, timetable adherences, and so forthof multitrains operating on the same rail line or network,research on minimizing their total TEC, especially in recentyears,makesmuch effort to get solutions from the perspectiveof systematic operation, as explained in Figure 5. This kindof research is able to reasonably contribute to decreasing theoverall TEC intensity of a rail transport system.

Kraay et al. [63] and Higgins et al. [64] have tried tooptimize the train operation diagram by utilizing mathe-matical programming to adjust the speeds and priorities oftrains in an integrated manner for reducing their delays andtotal TEC. In view of the transport characteristics of thetrains operating on the railway network of China in reality,Peng et al. [65] establish a schedule optimization modelwith the multiobjectives of optimal or near optimum totalTTE and TEC of trains. The practicability and adaptabilityof the proposed model for complex network structure areguaranteed, and a decomposition algorithm is put forwardto solve this model. In consideration of both structurecomplexity of a rail network and difference of passengercarrying capacities of different types of trains, Ghoseiri etal. [66] propose a new train scheduling model with moreobjectives besides decreasing the total TEC andTTEof trains,and the Pareto optimality [67] is suggested to be used to solvethe model.

To reduce the TEC of a rail transport system andmeanwhile maintain a certain service quality, dwell times oftrains at stations or/and their interstop TTEs need dynamic

TEC intensity

Service qualityInteractions

System operation

Timetable adherencesRegenerative braking

Etc.

Figure 5: Rationally systematic operation for TEC saving.

adjustments to match time-varying travel demand.The dwelltime adjustment is commonly utilized owing to its simplicityin practice [68]. For instance, a Heuristic-based train oper-ation controller has been developed by Sanso and Girard[69] to reduce the power demand of a metro system inpeak hours by introducing optimal dwell times for trains atsuccessive stations. Because adjusting the interstop TTEs of atrain with coasting controls can easily realize the exchangebetween its TEC and TTE for a whole trip [70], Wongand Ho [68] apply DP based on an event-based modelto rationally optimize the interstop TTEs of metro trainsaccording to dynamically changed passenger travel demandover a regional level especially in off-peak hours for energysaving under the premise of ensuring a certain level of thetransport service.

If trains running on different rail lines are approachingthe same transport section, the green wave strategy [71]for reducing their brakes and stops before entering theconflicting section is able to effectively decrease the TECs ofthese trains for their transports [72]. In order to make trainsrun through a conflicting transport section in the smoothestway, Bai et al. [73] apply golden section search method[74] to determine the most suitable speeds of trains whenthey are approaching a conflicting section. In comparisonwith the maximum traction strategy with the principleof improving the speed of a train before running into aconflicting transport section to its target speeds as much aspossible, the proposed approach may save about 6% of theTEC of such a train. Moreover, Fu et al. [75] make use of aGA with changeable lengths of chromosomes to analyze theinterimpacts of different trains passing through the same stopfrom different directions at nearly the same time. Accordingto the analyzed inter-impacts, corresponding energy-savingtrain control rules are suggested.

In an integrated electricity supply viewpoint, Falvo et al.[76] study the energy-saving effect of regenerative braking[77] of metro trains running on a rail network. The utiliza-tions of the electricity provided by the traction electric powersupply system and the electric energy from the regenerativebraking of metro trains for the tractions of trains on thesame metro network are recorded in the simulation study ofthe transport operations of these trains. Regenerative brakingenables much energy saving by not only injecting part of thebraking kinetic energy of a train into the electrical grid to beconsumed by nearby powering trains or returned to the AC


transmission system through a reversible substation [78] butalso preventing the tunnel temperature rising in undergroundrailways [79] to minimize the energy consumption in airconditioning or ventilating.The average utilization rate of theregenerative electric power for train tractions decreases withincreasing the headways of multitrains even if all of them areequipped with electrical energy storage devices such as theflywheel system [80], and a train must activate the rheostaticbraking to burn some regenerated power when the catenaryvoltage has reached its top limit [78].The unused regenerativeelectric power for train traction is suggested to serve otherelectricity consumption systems, for example, electric vehiclecharging stations in urban area, especially when the headwaysof the trains are relatively big.

In consideration of the interactions of trains in theirfollowing transports on the same rail line, Ding et al.[81] establish a dynamics model to optimize the energy-saving driving controls of following trains influenced by thedynamically changes of rail transport signals, according totheir previously proposed simulation algorithm [82] for theenergy-saving control of a single train under its fixed traveltime constraint for a certain transport. The correspondingHeuristic algorithm is also proposed to estimate the dynamicsmodel. Based on this model, the computer-aided simula-tion system is developed to get the optimal TEC of trainfollowing operation under the automatic block system [16].This dynamics modeling research is capable of optimizingthe controls of following trains according to the signals toavoid unnecessary brakes for adjusting their speeds. As aresult, the TEC of such trains is decreased. Additionally,taking into account the energy saving effect of regenerativebraking, Acikbas and Soylemez [83] make use of artificialneural networks [84, 85] to develop the simulation systemtaking a GA to search for systematically rational coastingstart points of different trains. In a different manner, Yanget al. [86] first establish the penalty function to evaluatethe influence of the interactions of trains upon their TTEsand TECs. Afterwards, the weight values of the overall TTEand TEC for an integrated cost evaluation are obtained byspreadsheet analyses. At last, a GA is accordingly utilized toexplore the optimal control program in a systematic way.

Because of the uncertain delays of trains due to the inter-influence of their transports on the same rail line or network,Ding et al. [87] and Cucala et al. [88] combine coastingprogram optimization and transport schedule modeling [89,90] to minimize the TEC of multitrains. In the transportsimulation study, a GA is applied to explore energy-savingcoasting schemes of the trains and obtain the relationshipbetween the corresponding TEC and TTE of each scheme.In order to realize the least TEC for the trains resuming atransport time schedule in case of an uncertain delay, Ding etal. [87] distribute the slack travel time of the rail line to eachof its transport sections simply according to the previouslyobtained relationship between the TEC and the TTE mainlyin view of the track alignments in all the transport sectionsand the length of every stopspacing. In contrast, based on therelationship between the TEC and the TTE, the fuzzy linearprogramming [91] is utilized by Cucala et al. [88] to rationallymake use of the slack time of a rail line for the delays which

are modeled as fuzzy numbers and punctuality constraintsand capable of getting reasonable behavioral responses of thedrivers for the most effective decrease of the general TEC.Such an approach is also applied to some Spanish HSR linesto evaluate the energy-saving potential of its commercialtransport service.

In view of the TECs and carbon emissions of trains andthe TTEs of their passengers, Li et al. [92] recently developa green train scheduling model to optimize the transporttimetable of multitrains in the operation of a rail transportsystem for the reduction of the system cost. The TEC of eachtrain is determined according to the correlations between itsrunning resistance, speed, and TTE for a transport work.Thecarbon emission cost of all the trains is computed based ontheir allowed and actual emission volumes. The total TTE ofthe passengers is used to describe the travel demand in a cer-tain time period. This multiobjective nonlinear optimizationproblem is solved with the utilization of fuzzy mathematicalprogramming [91]. It is proved that taking into account theTTE of all the passengers of multitrains operating on thesame rail line or network, the newly developed model is ableto effectively decrease the TEC and carbon emission of thesystematic transport operation.

7. Summary

There is no doubt that as to the driving strategy optimizationon the microlevel, a rational control program, especially avalid coasting scheme, of a train enables an evident decreaseof its TEC for the same transport work. As a result, manystudies have been made to explore reasonable approachesthat are able to efficiently determine in particular the optimalcoasting start points of a train in detail for its certain trip toultimately realize dynamic optimization of its driving tacticson board for energy saving. Meanwhile, impacts of varioustypes of track alignments on the TEC intensity of a trainneed adequate analyses in its operating control optimization.Moreover, it is also clarified on the mesolevel that the ame-liorative attributes such as reduced weight, improved poweroutput efficiency, decreased accelerating-braking frequency,and so forth of a train all play important roles in decreasingthe TEC of the whole transport operation process of the train.Furthermore, from the perspective of the transport organiza-tion on the macrolevel, the interinfluence of multitrains intheir coordinate operations requires systematic solutions toradically decrease the overall TEC intensity of a rail transportsystem in an effective manner.

In view of both instability of a system when its behaviorinfluenced by multifactors changes seriously [93] and thesystematic characteristics of the transport operation of a railline or network, dynamically systematical optimization ofthe control programs of trains with reasonably amendatoryattributes in sound responses to track alignments is necessaryto be further studied in an integrated way as much as possiblein future research on traction energy saving of rail transport.This is also much beneficial to the improvement of railtransport safety [94, 95]. Of course, utilizing new energy, for


example, hydrogen against electrification [96, 97], deservesour continual effort as well.

Acknowledgments

This research is supported by the National Natural ScienceFoundation of China (71201006 and 71131001), the NationalBasic Research Program of China (2012CB725406) and theFundamental Research Funds for the Central Universities(2011JBM251).

References

[1] P. C.M. Leung and E.W.M. Lee, “Estimation of electrical powerconsumption in subway station design by intelligent approach,”Applied Energy, vol. 101, pp. 634–643, 2013.

[2] C. Winston and V. Maheshri, “On the social desirability ofurban rail transit systems,” Journal of Urban Economics, vol. 62,no. 2, pp. 362–382, 2007.

[3] J. Hickman, D.Hassel, R. Joumard, Z. Samaras, and S. Sorenson,Methodology for Calculating Transport Emissions and EnergyConsumption, Office for Official Publications of the EuropeanCommunities, Luxembourg, Belgium, 1999.

[4] X. Feng, “Energy cost of high-speed railway trains in theeffect of utilization ratios of passenger seats,” in Proceedingsof the International Conference on Information, Services andManagement Engineering (ISME ’11), vol. 1, pp. 133–136, Beijing,China, December 2011.

[5] E. Lindgreen and S. C. Sorenson, Simulation of Energy Con-sumption and Emissions from Rail Traffic Evaluation, TechnicalUniversity of Denmark, Lyngby, Denmark, 2005.

[6] P. Lukaszewicz, Energy consumption and running time for trains:modelling of running resistance and driver behavior based onfull scale testing [Ph.D. thesis], Royal Institute of Technology,Stockholm, Sweden, 2001.

[7] A. R. Miller, J. Peters, B. E. Smith, and O. A. Velev, “Analysis offuel cell hybrid locomotives,” Journal of Power Sources, vol. 157,no. 2, pp. 855–861, 2006.

[8] H. Wang, H. Hao, X. Li, K. Zhang, and M. Ouyang, “Perfor-mance of Euro III common rail heavy duty diesel engine fueledwith gas to liquid,”Applied Energy, vol. 86, no. 10, pp. 2257–2261,2009.

[9] E. V. Hoyt and R. R. Levary, “Assessing the effects of severalvariables on freight train fuel consumption and performanceusing a train performance simulator,” Transportation ResearchA, vol. 24, no. 2, pp. 99–112, 1990.

[10] R. S. Raghunathan,H.D. Kim, andT. Setoguchi, “Aerodynamicsof high-speed railway train,” Progress in Aerospace Sciences, vol.38, no. 6-7, pp. 469–514, 2002.

[11] H. I. Andrews, Railway Traction: The Principles of Mechanicaland Electrical Railway Traction, vol. 5 of Studies in MechanicalEngineering, Elsevier Scientific, New York, NY, USA, 1986.

[12] C. Chandra and M. M. Aqarwal, Railway Engineering, OxfordHigher Education, New York, NY, USA, 2008.

[13] A. Chui, K. K. Li, and P. K. Lau, “Traction energy managementin KCR,” in Proceedings of the 2nd International Conference onAdvances in Power System Control, Operation & Management,vol. 1, pp. 202–208, December 1993.

[14] C. S. Chang and S. S. Sim, “Optimising train movementsthrough coast control using genetic algorithms,” IEE Pro-ceedings—Electric Power Application, vol. 144, no. 1, pp. 65–73,1997.

[15] J. S. Arora, “Genetic algorithms for optimum design,” inIntroduction to Optimum Design, J. S. Arora, Ed., pp. 643–655,Academic Press, Waltham, UK, 3rd edition, 2012.

[16] B. Solomon, Railroad Signaling, Voyageur Press, Minneapolis,Minn, USA, 2010.

[17] T. H. Cormen, C. E. Leiserson, R. L. Rivest, and C. Stein,Introduction to Algorithms, MIT Press, Cambridge, Mass, USA,3rd edition, 2009.

[18] K. K. Wong and T. K. Ho, “Coast control for mass rapid transitrailways with searching methods,” IEE Proceedings: ElectricPower Applications, vol. 151, no. 3, pp. 365–375, 2004.

[19] H. Hwang, “Control strategy for optimal compromise betweentrip time and energy consumption in a high-speed railway,”IEEE Transactions on Systems, Man, and Cybernetics A, vol. 28,no. 6, pp. 791–802, 1998.

[20] F. Hppner, F. Klawonn, K. Kruse, and T. Runkler, Fuzzy ClusterAnalysis: Methods for Classification, Data Analysis and ImageRecognition, John Wiley & Sons, Chichester, UK, 1999.

[21] M. S. Yang, “A survey of fuzzy clustering,” Mathematical andComputer Modelling, vol. 18, no. 11, pp. 1–16, 1993.

[22] R. Liu and I. M. Golovitcher, “Energy-efficient operation of railvehicles,” Transportation Research A, vol. 37, no. 10, pp. 917–932,2003.

[23] Y. Bai, B.Mao, Y.Ding, F. Zhou, andW. Jia, “An onboard optimalcontrol system for freight trains,” in Proceedings of the 6thInternational Conference on Traffic and Transportation StudiesCongress, pp. 770–783, Nanning, China, August 2008.

[24] A. Ioffe andV. Tikhomirov,Theory of External Problems, Nauka,Moscow, Russia, 1974.

[25] R. E. Showalter, “The maximum principle,” in Introductionto Holomorphy, J. A. Barroso, Ed., vol. 106 of North-HollandMathematics Studies, chapter 10, pp. 115–117, Elsevier ScienceB.V., Amsterdam, The Netherlands, 1985.

[26] Y. V. Bocharnikov, A. M. Tobias, C. Roberts, S. Hillmansen,and C. J. Goodman, “Optimal driving strategy for tractionenergy saving on DC suburban railways,” IET Electric PowerApplications, vol. 1, no. 5, pp. 675–682, 2007.

[27] T. R. Kane and D. A. Levinson, Dynamics: Theory and Applica-tions, McGraw-Hill, New York, NY, USA, 1985.

[28] J. N. Mordeson, “Fuzzy mathematics,” in Foundations of ImageUnderstanding, L. S. Davis, Ed., pp. 95–126, Springer Science,Business Media, New York, NY, USA, 2001.

[29] K. Kim and S. I. Chien, “Simulation-based analysis of train con-trols under various track alignments,” Journal of TransportationEngineering, vol. 136, no. 11, pp. 937–948, 2010.

[30] D. Bertsimas and J. Tsitsiklis, “Simulated annealing,” StatisticalScience, vol. 8, no. 1, pp. 10–15, 1993.

[31] K. Kim and S. I. Chien, “Optimal train operation for minimumenergy consumption considering track alignment, speed limit,and schedule adherence,” Journal of Transportation Engineering,vol. 137, no. 9, pp. 665–674, 2011.

[32] G. F. Ignacio and G. A. Alberto, “Can high-speed trains runfaster and reduce energy consumption,” Procedia—Social andBehavioral Sciences, vol. 48, pp. 827–837, 2012.

[33] P.Howlett, “Theoptimal control of a train,”Annals of OperationsResearch, vol. 98, no. 1–4, pp. 65–87, 2000.


[34] P. G. Howlett, P. J. Pudney, and X. Vu, “Local energy minimiza-tion in optimal train control,” Automatica, vol. 45, no. 11, pp.2692–2698, 2009.

[35] H. Liu, B. Mao, Y. Ding, W. Jia, and S. Lai, “Train energy-saving scheme with evaluation in urban mass transit systems,”Journal of Transportation Systems Engineering and InformationTechnology, vol. 7, no. 5, pp. 68–73, 2007.

[36] H. H. Hoang, M. P. Polis, and A. Haurie, “Reducing energyconsumption through trajectory optimization for a metronetwork,” IEEE Transactions on Automatic Control, vol. 20, no.5, pp. 590–595, 1975.

[37] S. Edelkamp and S. Schrdl, Heuristic Search: Theory and Appli-cations, Morgon Kaulfmann, Waltham, Mass, USA, 2011.

[38] E. V.Denardo,Dynamic Programming:Models andApplications,Dover, New York, NY, USA, 2003.

[39] P. Firpo and S. Savio, “Optimized train running curve forelectrical energy saving in autotransformer supplied AC railwaysystems,” in Proceedings of the International Conference onElectric Railways in aUnited Europe, pp. 23–27, Amsterdam,TheNetherlands, March 1995.

[40] S. E. Lyshevski, Electromechanical Systems, Electric Machines,and Applied Mechatronics, CRC Press, Boca Raton, Fla, USA,1999.

[41] I. Maros, Computational Techniques of the Simplex Method,Kluwer Academic, Dordrecht, The Netherlands, 2002.

[42] D. A. Coley, An Introduction to Genetic Algorithms for Scientistsand Engineers, World Scientific, Singapore, 1999.

[43] A. W. Johnson and S. H. Jacobson, “On the convergence ofgeneralized hill climbing algorithms,” Discrete Applied Mathe-matics, vol. 119, no. 1-2, pp. 37–57, 2002.

[44] D. N. Kim and P. M. Schonfeld, “Benefits of dipped verticalalignments for rail transit routes,” Journal of TransportationEngineering, vol. 123, no. 1, pp. 20–27, 1997.

[45] B. R. Ke, M. C. Chen, and C. L. Lin, “Block-layout designusing maxmin ant system for saving energy on mass rapidtransit systems,” IEEE Transactions on Intelligent TransportationSystems, vol. 10, no. 2, pp. 226–235, 2009.

[46] T. Stutzle and H. H. Hoos, “MAX-MIN ant system,” FutureGeneration Computer Systems, vol. 16, no. 8, pp. 889–914, 2000.

[47] B. R. Ke, C. L. Lin, and C. W. Lai, “Optimization of train-speedtrajectory and control for mass rapid transit systems,” ControlEngineering Practice, vol. 19, no. 7, pp. 675–687, 2011.

[48] B. Liu, “Fuzzy process, hybrid process and uncertain process,”Journal of Uncertain Systems, vol. 2, no. 1, pp. 3–16, 2008.

[49] K.M. Passino and S. Yurkovich,FuzzyControl, Addison-Wesley,Longman, Menlo Park, Calif, USA, 1998.

[50] K. P. Kokotovic and S. G. Singh, “Minimum-energy control ofa traction motor,” IEEE Transactions on Automatic Control, vol.17, no. 1, pp. 92–95, 1972.

[51] A. Blokhuis, “Hamiltonian systems,” in Methods of DifferentialGeometry in Analytical Mechanics, M. Len and P. R. Rodrigues,Eds., vol. 158 of North-Holland Mathematics Studies, chapter6, pp. 263–299, Elsevier Science B.V., Amsterdam, The Nether-lands, 1989.

[52] R. H. Martin and M. Pierre, “The adjoint operator method,”in Computational Methods in Engineering Boundary ValueProblems, T. Y. Na, Ed., vol. 145 of Mathematics in Science andEngineering, chapter 4, pp. 52–69, Academic Press, New York,NY, USA, 1979.

[53] A. Miele, “Extremization of linear integrals by green’s theorem,”in Optimization Techniques with Applications to Aerospace Sys-tems, G. Leitmann, Ed., vol. 5 of Mathematics in Science andEngineering, pp. 69–98, Academic Press, New York, NY, USA,1962.

[54] IFEU (Institut fur Energieund Umweltforschung HeidelbergGmbH), Energy Savings by Light-Weighting (Final Report),International Aluminium Institute (IAI), Heidelberg, Germany,2003.

[55] M. Chou and X. Xia, “Optimal cruise control of heavy-haultrains equipped with electronically controlled pneumatic brakesystems,” Control Engineering Practice, vol. 15, no. 5, pp. 511–519,2007.

[56] S. Lu, S. Hillmansen, and C. Roberts, “A power-managementstrategy for multiple-unit railroad vehicles,” IEEE Transactionson Vehicular Technology, vol. 60, no. 2, pp. 406–420, 2011.

[57] Y. Zhu, Y. Chen, G. Tian, H. Wu, and Q. Chen, “A four-stepmethod to design an energy management strategy for hybridvehicles,” in Proceedings of the American Control Conference(AAC ’04), pp. 156–161, Boston, Mass, USA, July 2004.

[58] X. Feng, “Optimization of target speeds of high-speed rail-way trains for traction energy saving and transport efficiencyimprovement,”Energy Policy, vol. 39, no. 12, pp. 7658–7665, 2011.

[59] X. Feng, B. Mao, X. Feng, and J. Feng, “Study on the maximumoperation speeds of metro trains for energy saving as well astransport efficiency improvement,” Energy, vol. 36, no. 11, pp.6577–6582, 2011.

[60] R. A. Uher, N. Sathi, and A. Sathi, “Traction energy costreduction of theWMATAmetro rail system,” IEEE Transactionson Industry Applications, vol. IA-20, no. 3, pp. 472–483, 1984.

[61] X. Feng, J. Feng, K. Wu, H. Liu, and Q. Sun, “Evaluating targetspeeds of passenger trains in China for energy saving in theeffect of different formation scales and traction capacities,”International Journal of Electrical Power andEnergy Systems, vol.42, no. 1, pp. 621–626, 2012.

[62] I. Lopez, J. Rodrıguez, J. M. Buron, and A. Garcıa, “A method-ology for evaluating environmental impacts of railway freighttransportation policies,” Energy Policy, vol. 37, no. 12, pp. 5393–5398, 2009.

[63] D. Kraay, P. T. Harker, and B. Chen, “Optimal pacing of trains infreight railroads: model formulation and solution,” OperationsResearch, vol. 39, no. 1, pp. 82–99, 1991.

[64] A. Higgins, E. Kozan, and L. Ferreira, “Optimal scheduling oftrains on a single line track,” Transportation Research B, vol. 30,no. 2, pp. 147–161, 1996.

[65] Q. Peng, S. Zhu, and P. Wang, “Study on the strategy of trainoperation adjustment on high speed railway,” Journal of theChina Railway Society, vol. 23, no. 1, pp. 1–8, 2001.

[66] K. Ghoseiri, F. Szidarovszky, and M. J. Asgharpour, “A multi-objective train scheduling model and solution,” TransportationResearch B, vol. 38, no. 10, pp. 927–952, 2004.

[67] M. Brio, G. M. Webb, and A. R. Zakharian, “Pareto optimal-ity,” in The Vector-Valued Maximin, V. I. Zhukovskiy and M.E. Salukvadze, Eds., vol. 193 of Mathematics in Science andEngineering, chapter 3, pp. 81–129, Academic Press, New York,NY, USA, 1994.

[68] K. K.Wong and T. K. Ho, “Dwell-time and run-time control forDCmass rapid transit railways,” IET Electric Power Applications,vol. 1, no. 6, pp. 956–966, 2007.


[69] B. Sanso and P. Girard, “Train scheduling desynchronizationand power peak optimization in a subway system,” in Proceed-ings of the IEEE/ASME Joint Railroad Conference, pp. 75–78,Baltimore, Md, USA, April 1995.

[70] K. K. Wong and T. K. Ho, “Dynamic coast control of trainmovement with genetic algorithm,” International Journal ofSystems Science, vol. 35, no. 13-14, pp. 835–846, 2004.

[71] T. Albrecht, “The influence of anticipating train driving on thedispatching process in railway conflict situations,”Networks andSpatial Economics, vol. 9, no. 1, pp. 85–101, 2009.

[72] F. Corman, A. D’Ariano, D. Pacciarelli, and M. Pranzo, “Eval-uation of green wave policy in real-time railway traffic man-agement,” Transportation Research C, vol. 17, no. 6, pp. 607–616,2009.

[73] Y. Bai, T. O. Ho, and B. Mao, “Train control to reduce delaysupon service disturbances at railway junctions,” Journal ofTransportation Systems Engineering and Information Technol-ogy, vol. 11, no. 5, pp. 114–122, 2011.

[74] W. J. Braun and D. J. Murdoch, A First Course in StatisticalProgramming with R, Cambridge University Press, Cambridge,UK, 2007.

[75] Y. Fu, Z. Gao, and K. Li, “Optimizationmethod of energy savingtrain operation for railway network,” Journal of TransportationSystems Engineering and Information Technology, vol. 9, no. 4,pp. 90–96, 2009.

[76] M. C. Falvo, R. Lamedica, R. Bartoni, and G. Maranzano,“Energy management in metro-transit systems: an innovativeproposal toward an integrated and sustainable urban mobilitysystem including plug-in electric vehicles,” Electric Power Sys-tems Research, vol. 81, no. 12, pp. 2127–2138, 2011.

[77] S. Robertson and J. Markham, Regenerative Braking Story,Venture Publications, Glossop, UK, 2007.

[78] M. Coto, P. Arboleya, and C. Gonzalez-Moran, “Optimizationapproach to unified AC/DC power flow applied to tractionsystemswith catenary voltage constraints,” International Journalof Electrical Power and Energy Systems, vol. 53, no. 1, pp. 434–441, 2013.

[79] T. Suzuki, “DC power-supply system with inverting substationsfor traction systems using regenerative brakes,” IEE ProceedingsB, vol. 129, no. 1, pp. 18–26, 1982.

[80] M. B. Richardson, “Flywheel energy storage system for tractionapplications,” in Proceedings of the International Conference onPower Electronics,Machines andDrives, Conference Publicationno. 487, pp. 275–279, Bath, UK, April 2002.

[81] Y. Ding, F. Zhou, Y. Bai, and H. Liu, “Simulation algorithmfor energy-efficient train following operation under automatic-block system,” Journal of System Simulation, vol. 21, no. 15, pp.4593–4597, 2009.

[82] Y. Ding, B. Mao, H. Liu, T. Wang, and X. Zhang, “Algorithm forenergy-efficient train operation simulation with fixed runningtime,” Journal of System Simulation, vol. 16, no. 10, pp. 2241–2244, 2004.

[83] S. Acikbas and M. T. Soylemez, “Coasting point optimisationfor mass rail transit lines using artificial neural networks andgenetic algorithms,” IET Electric Power Applications, vol. 2, no.3, pp. 172–182, 2008.

[84] D. Anderson and G. McNeill, Artificial Neural Networks Tech-nology: A DACS State-of-the-Art Report, Kaman Sciences Cor-poration, New York, NY, USA, 1992.

[85] D. J. Livingstone, Artificial Neural Networks: Methods andApplications, Humana Press, Totowa, NJ, USA, 2008.

[86] L. Yang, K. Li, Z. Gao, and X. Li, “Optimizing trains movementon a railway network,” Omega, vol. 40, no. 5, pp. 619–633, 2012.

[87] Y. Ding, H. Liu, Y. Bai, and F. Zhou, “A two-level optimiza-tion model and algorithm for energy-efficient urban trainoperation,” Journal of Transportation Systems Engineering andInformation Technology, vol. 11, no. 1, pp. 96–101, 2011.

[88] A. P. Cucala, A. Fernandez, C. Sicre, andM.Domınguez, “Fuzzyoptimal schedule of high speed train operation to minimizeenergy consumption with uncertain delays and driver’s behav-ioral response,” Engineering Applications of Artificial Intelli-gence, vol. 25, no. 8, pp. 1548–1557, 2012.

[89] M. Hickman, P. Mirchandani, and S. Vo, Computer-AidedSystems in Public Transport, Springer, Berlin, Germany, 2008.

[90] M. Salicru, C. Fleurent, and J. M. Armengol, “Timetable-based operation in urban transport: run-time optimisationand improvements in the operating process,” TransportationResearch A, vol. 45, no. 8, pp. 721–740, 2011.

[91] Y. J. Lai and C. L. Hwang, Fuzzy Mathematical Programming:Methods and Applications, Springer, New York, NY, USA, 1995.

[92] X. Li, D. Wang, K. Li, and Z. Gao, “A green train schedul-ing model and fuzzy multi-objective optimization algorithm,”Applied Mathematical Modelling, vol. 37, no. 4, pp. 2063–2073,2013.

[93] W.Wang, X. Jiang, S. Xia, and Q. Cao, “Incident tree model andincident tree analysis method for quantified risk assessment: anin-depth accident study in traffic operation,” Safety Science, vol.48, no. 10, pp. 1248–1262, 2010.

[94] M. Nejlaoui, A. Houidi, Z. Affi, and L. Romdhane, “Multiobjec-tive robust design optimization of rail vehicle moving in shortradius curved tracks based on the safety and comfort criteria,”Simulation Modelling Practice and Theory, vol. 30, pp. 21–34,2013.

[95] E. Matsika, S. Ricci, P. Mortimer, N. Georgiev, and C. O’Neill,“Rail vehicles, environment, safety and security,” Research inTransportation Economics, vol. 41, no. 1, pp. 43–58, 2013.

[96] G. D. Marin, G. F. Naterer, and K. Gabriel, “Rail transportationby hydrogen versus electrification—case study for OntarioCanada—I: propulsion and storage,” International Journal ofHydrogen Energy, vol. 35, no. 12, pp. 6084–6096, 2010.

[97] G. D. Marin, G. F. Naterer, and K. Gabriel, “Rail transportationby hydrogen versus electrification—case study for Ontario,Canada—II: energy supply and distribution,” International Jour-nal of Hydrogen Energy, vol. 35, no. 12, pp. 6097–6107, 2010.

Submit your manuscripts athttp://www.hindawi.com

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

MathematicsJournal of


Mathematical Problems in Engineering

Hindawi Publishing Corporationhttp://www.hindawi.com

Differential EquationsInternational Journal of

Volume 2014

Applied MathematicsJournal of


Probability and StatisticsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Journal of


Mathematical PhysicsAdvances in

Complex AnalysisJournal of


OptimizationJournal of


CombinatoricsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

International Journal of


Operations ResearchAdvances in

Journal of


Function Spaces

Abstract and Applied AnalysisHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

International Journal of Mathematics and Mathematical Sciences


The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014


Algebra

Discrete Dynamics in Nature and Society



Decision SciencesAdvances in

Discrete MathematicsJournal of

Hindawi Publishing Corporationhttp://www.hindawi.com

Volume 2014 Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Stochastic AnalysisInternational Journal of

review article a review study on traction energy saving of...

Documents